michael wade (the beetle guy) and felix breden worked up some mathematical models of the possible frequencies of “genes for altruism” in several different types of inbreeding populations.
they took different degrees of inbreeding/outbreeding …
- mating with oneself or a clone (100% related)
- mating with a sibling (50% related)
- mating with a more distant relative (20% related, i.e. not quite half-siblings)
- no inbreeding at all (0% related)
… and via wizardry (i.e. advanced algebra) they worked out how “genes for altruism” would fare in each of these populations over the course of many generations. in other words, would altruism genes become more frequent or not in these various populations?
they factored in different parameters such as whether the gene(s) (alleles) in question were dominant or recessive, and whether the selection pressures on the alleles were weak or strong. weak selection apparently refers to those cases in which one “phenotype is slightly advantageous over another.” presumably strong selection means the opposite.
here’s what wade and breden found:
under weak selection — instances in which the altruism alleles only confer slight advantages to those who have them (top two graphs) — the altruism alleles really only increase in any significant way when the individuals self-mate (or mate with clones) or mate with their full-siblings. there’s some increase in altruism alleles in populations where mating occurs between individuals who are almost half-siblings and the alleles are dominant, but that increase really doesn’t become apparent until after several hundred generations of inbreeding.
under strong selection (lower two graphs), again the altruism alleles increase in frequency the most when the individuals self-mate or mate with full-sibs. however, there is also a marked increase in populations where mating occurs between individuals who are almost half-sibs AND the alleles are dominant. in fact, the slope really takes off after just fifty generations or so (solid line, bottom graph).
the authors conclude that: “Increasing the level of inbreeding can greatly increase the rate of change of gene frequency of the altruistic allele.”
i’m interested in the evolution of altruism in humans, though, and not many humans mate with themselves (yet) or even their full-siblings. what’s more common, as we all know by now, is cousin marriage.
mating with your first-cousin in a population where inbreeding doesn’t normally occur means your relatedness to your cousin is probably around 12.5%, much lower than the lowest inbreeding rate that wade and breden looked at (20%). however, in populations where inbreeding is frequent and regular, the coefficients of relatedness are much higher — for instance, some (many?) pakistani and saudi cousins have a coefficient of relatedness of around 22% (11% coefficient of inbreeding x 2). that’s pretty much the same as the lowest degree of inbreeding that wade and breden looked at.
i think it’s apparent by looking at human behavior that inbreeding affects the frequencies of altruism alleles in different human populations, but since we don’t even know what those alleles are yet, this hasn’t been proven one hundred percent. if wade and breden did their sums right, then my guess is that (at least some) altruism alleles in humans must be dominant and must confer a good deal of advantage to those who have them. in other words, if we could graph the frequencies of altruism alleles in humans who marry their cousins regularly over time, i think they would look something like the bottom graph above, although perhaps with a trajectory that wasn’t quite so sharp (since in no population does cousin marriage happen one hundred percent of the time in every generation).
the arabs, for example, have been marrying their first-cousins (often double-first-cousins) since at least mohammed’s days, or something like 1400 years ago. if we take a very conservative generation length as twenty-five years, that’s roughly 56 generations of inbreeding up to the present. at least. plenty of time, according to wade and breden, for altruism alleles to increase in that population — provided the alleles are dominant and the selection is strong.
a couple of other things to keep in mind: 1) like genes for height or intelligence, there are probably many genes for altruism, so we have to imagine some sort of cumulative effect of many genes on human behavior, i.e. we’d have to draw many charts to map the frequencies of many genes; 2) individuals in a population might share lots of alleles for reasons other than recent inbreeding, such as a population’s ancestors having gone through a bottleneck at some point in the past. you’d think that that could also contribute to the number of shared altruism alleles in a population.
update 04/30: see also more on inbreeding and the evolution of altruistic behavior
update 05/30: see also inbreeding and the evolution of altruistic behavior ii
(note: comments do not require an email. citizens against altruism!)